INDEX

Explanations

scientific and analytical terms related to hypotheses, theories, assumptions, and arguments

oai_token-act-pair · gpt-3.5-turbo

assertions or hypotheses that relate to social theories and misconceptions

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

aths

-0.75

iencies

-0.72

 Chains

-0.72

quished

-0.72

 downtime

-0.71

 curfew

-0.70

orers

-0.70

 Nanto

-0.70

dule

-0.68

 Kens

-0.67

POSITIVE LOGITS

 unfounded

1.15

 baseless

1.13

 valid

1.09

 refuted

1.01

 echoed

1.01

 nonsense

1.00

 false

1.00

 untrue

0.99

 debunked

0.97

 incorrect

0.97

Activations Density 0.460%

scientific and analytical terms related to hypotheses, theories, assumptions, and arguments

assertions or hypotheses that relate to social theories and misconceptions

No Comments

No Known Activations