INDEX

Explanations

complex strings of characters at various levels of activation

oai_token-act-pair · gpt-3.5-turbo

alphanumeric sequences and upper-case letters

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

itsch

-0.82

DonaldTrump

-0.78

à©

-0.72

taboola

-0.70

illary

-0.70

âĸ¬

-0.69

 GOODMAN

-0.69

ãĥ¯ãĥ³

-0.68

idays

-0.68

istries

-0.67

POSITIVE LOGITS

fy

0.81

ZX

0.73

qq

0.72

\">

0.70

Bs

0.68

XM

0.67

 Shib

0.67

dn

0.66

dq

0.65

Nab

0.65

Activations Density 0.074%

complex strings of characters at various levels of activation

alphanumeric sequences and upper-case letters

No Comments

No Known Activations