INDEX

Explanations

phrases related to actions and tasks that involve following a specific set of instructions or steps

oai_token-act-pair · gpt-3.5-turbo

phrases related to collaborative actions and decision-making processes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

imov

-0.57

\":

-0.56

nud

-0.52

 Rossi

-0.52

gov

-0.51

 Morty

-0.50

 Thur

-0.49

netflix

-0.49

................................................................

-0.48

 Polk

-0.47

POSITIVE LOGITS

 desired

0.62

rue

0.58

odox

0.58

orage

0.55

eport

0.54

 favourable

0.53

 uninterrupted

0.52

ones

0.52

 safely

0.52

 meaningful

0.50

Activations Density 1.312%

phrases related to actions and tasks that involve following a specific set of instructions or steps

phrases related to collaborative actions and decision-making processes

No Comments

No Known Activations