INDEX

Explanations

phrases related to agreements or some form of official directives

oai_token-act-pair · gpt-3.5-turbo

terms related to agreements and calls to action

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

cale

-0.77

ecause

-0.74

vae

-0.69

ndum

-0.67

MORE

-0.66

agine

-0.65

poon

-0.64

ourke

-0.62

Interested

-0.62

stasy

-0.62

POSITIVE LOGITS

 itself

0.82

iest

0.78

ariat

0.75

 consists

0.71

 revolves

0.70

 consisted

0.70

 factor

0.67

ultimate

0.66

 interval

0.66

 seemed

0.66

Activations Density 0.437%

phrases related to agreements or some form of official directives

terms related to agreements and calls to action

No Comments

No Known Activations