INDEX

Explanations

locations or date references

oai_token-act-pair · gpt-3.5-turbo

specific references to locations or geographic features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

thood

-0.69

omon

-0.67

aea

-0.65

aukee

-0.64

isite

-0.62

ento

-0.62

olars

-0.61

elist

-0.60

agne

-0.60

acia

-0.59

POSITIVE LOGITS

insk

0.59

 Worst

0.54

 Thing

0.53

ortunate

0.52

 parts

0.52

Dialog

0.51

Dex

0.50

 best

0.50

md

0.49

ines

0.49

Activations Density 0.200%

locations or date references

specific references to locations or geographic features

No Comments

No Known Activations