INDEX

Explanations

mentions of specific names related to media personalities

oai_token-act-pair · gpt-3.5-turbo

prominent names and references related to media and politics

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

orative

-0.77

llular

-0.76

igm

-0.75

eele

-0.70

acular

-0.70

afort

-0.69

ision

-0.68

uality

-0.68

pmwiki

-0.66

zyk

-0.66

POSITIVE LOGITS

mong

0.85

IGH

0.84

masters

0.79

sworth

0.74

wig

0.72

ollow

0.71

Rid

0.67

ISTORY

0.67

idays

0.67

0.66

Activations Density 0.116%

mentions of specific names related to media personalities

prominent names and references related to media and politics

No Comments

No Known Activations