INDEX

Explanations

references to pointing or attribution of blame

oai_token-act-pair · gpt-3.5-turbo

phrases or expressions pointing to specific subjects or claims

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

lance

-0.80

soever

-0.74

hop

-0.74

stress

-0.73

proxy

-0.73

unte

-0.72

itialized

-0.70

mind

-0.67

DH

-0.67

 operates

-0.66

POSITIVE LOGITS

WARD

0.68

TextColor

0.68

odan

0.67

othy

0.67

 finger

0.67

 Genocide

0.64

obin

0.64

 victory

0.62

 evidence

0.61

 Victory

0.61

Activations Density 0.075%

references to pointing or attribution of blame

phrases or expressions pointing to specific subjects or claims

No Comments

No Known Activations