INDEX

Explanations

specific combinations of characters that signify code or commands

oai_token-act-pair · gpt-3.5-turbo

keywords related to significant statements or declarations

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

quartered

-0.59

ypes

-0.57

doms

-0.56

uits

-0.53

eno

-0.53

ricular

-0.52

unker

-0.52

leground

-0.51

nces

-0.51

cipl

-0.51

POSITIVE LOGITS

 itself

0.65

ciation

0.54

 afore

0.54

 Cortana

0.53

liest

0.47

Completed

0.46

 Hoover

0.46

 minus

0.45

LV

0.44

 thereof

0.44

Activations Density 1.902%

specific combinations of characters that signify code or commands

keywords related to significant statements or declarations

No Comments

No Known Activations