INDEX

Explanations

names of cities/towns

oai_token-act-pair · gpt-3.5-turbo

words related to anatomy or geographical locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bler

-0.74

ples

-0.72

TOR

-0.67

tackle

-0.66

blers

-0.65

bly

-0.62

Knight

-0.61

 chew

-0.60

yles

-0.58

 Oath

-0.57

POSITIVE LOGITS

ifa

0.82

ashtra

0.77

¬

0.73

amina

0.73

ische

0.67

®

0.66

uan

0.64

Ã®

0.64

 Paulo

0.63

ienne

0.63

Activations Density 0.123%

names of cities/towns

words related to anatomy or geographical locations

No Comments

No Known Activations