INDEX

Explanations

proper nouns or names in text

oai_token-act-pair · gpt-3.5-turbo

specific names and titles related to people, places, and organizations

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

astical

-0.74

 Gaul

-0.71

quartered

-0.67

rious

-0.66

ullah

-0.65

nikov

-0.64

pard

-0.61

åĬ

-0.61

ophob

-0.61

tics

-0.60

POSITIVE LOGITS

tracks

0.72

ITION

0.72

 debit

0.64

ĨĴ

0.63

wagon

0.62

ERT

0.61

sor

0.60

metics

0.59

 cert

0.59

DEV

0.59

Activations Density 0.360%

proper nouns or names in text

specific names and titles related to people, places, and organizations

No Comments

No Known Activations