INDEX

Explanations

terms related to responses and reactions

oai_token-act-pair · gpt-3.5-turbo

the presence of the term "respond" in various forms, indicating discussions around response or accountability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

BALL

-0.75

 Rwanda

-0.70

fare

-0.70

 AMERICA

-0.66

 devils

-0.63

 Muller

-0.63

 caveat

-0.62

 Maurit

-0.60

 caution

-0.60

 Staten

-0.59

POSITIVE LOGITS

onding

1.45

ibilities

1.23

awn

1.17

onds

1.13

ond

1.06

onder

1.06

ublic

1.03

ible

1.00

ons

0.93

itable

0.92

Activations Density 0.038%

terms related to responses and reactions

the presence of the term "respond" in various forms, indicating discussions around response or accountability

No Comments

No Known Activations