INDEX

Explanations

names of cities or locations

oai_token-act-pair · gpt-3.5-turbo

names and terms related to geography and places

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

ï¸ı

-0.66

 daddy

-0.61

 meaning

-0.60

Fun

-0.58

aminer

-0.58

 robotics

-0.56

 hint

-0.56

CTR

-0.55

WT

-0.55

MY

-0.55

POSITIVE LOGITS

atown

0.90

ufact

0.87

ouver

0.83

oba

0.81

ensis

0.74

atu

0.73

artney

0.72

querque

0.70

arie

0.70

hari

0.70

Activations Density 0.116%

names of cities or locations

names and terms related to geography and places

No Comments

No Known Activations