INDEX

Explanations

numeric values representing times

oai_token-act-pair · gpt-3.5-turbo

numerical values and time indicators

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

raltar

-0.72

idates

-0.68

rament

-0.67

omnia

-0.65

necess

-0.62

 governors

-0.60

pter

-0.60

unal

-0.59

anship

-0.59

upiter

-0.59

POSITIVE LOGITS

BSD

0.81

[+]

0.79

0.78

 WRITE

0.76

0.72

 =================================================================

0.67

[*]

0.66

#$

0.66

 Contrast

0.65

WAR

0.64

Activations Density 0.081%

numeric values representing times

numerical values and time indicators

No Comments

No Known Activations