INDEX

Explanations

words related to specific names or entities

oai_token-act-pair · gpt-3.5-turbo

names and terms related to individuals and possible false information or deceit

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ablishment

-0.93

 Zucker

-0.83

steen

-0.79

 Kart

-0.73

bered

-0.73

asury

-0.73

uality

-0.72

 Freeze

-0.71

eele

-0.71

utsu

-0.70

POSITIVE LOGITS

ouse

0.74

ĪĴ

0.74

Ĩ

0.72

 chemistry

0.71

sheet

0.70

oyd

0.69

 compe

0.69

ŃĶ

0.69

 seaf

0.68

 enthusi

0.67

Activations Density 0.016%

words related to specific names or entities

names and terms related to individuals and possible false information or deceit

No Comments

No Known Activations