INDEX

Explanations

statements or claims supported by evidence or lack of evidence

oai_token-act-pair · gpt-3.5-turbo

phrases related to evidence and corroboration of claims

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 actionGroup

-1.01

ktop

-0.89

zie

-0.82

ourning

-0.76

ophon

-0.71

 Lumpur

-0.70

unicip

-0.68

enfranch

-0.68

isure

-0.68

onding

-0.68

POSITIVE LOGITS

 assertions

1.48

 debunked

1.38

 claims

1.31

 falsehood

1.27

 untrue

1.24

 assertion

1.21

 baseless

1.21

 allegations

1.18

Claim

1.17

 factual

1.17

Activations Density 0.948%

statements or claims supported by evidence or lack of evidence

phrases related to evidence and corroboration of claims

No Comments

No Known Activations