INDEX

Explanations

contradictory statements in the text

oai_token-act-pair · gpt-3.5-turbo

instances of denial or negation in statements

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

IFrame

Link

Negative Logits

 paced

-0.66

abal

-0.65

 hurd

-0.64

FIG

-0.64

PID

-0.64

adelphia

-0.64

 afloat

-0.63

\-

-0.62

 comprom

-0.62

 enthusi

-0.60

POSITIVE LOGITS

sir

0.90

 wait

0.87

onsense

0.86

 Nope

0.83

 Absolutely

0.77

 Sorry

0.74

nor

0.74

 none

0.72

 thank

0.71

 Didn

0.70

Activations Density 0.079%

contradictory statements in the text

instances of denial or negation in statements

No Known Activations