INDEX

Explanations

countries or states in the format "In [location]," followed by a significant event or situation

oai_token-act-pair · gpt-3.5-turbo

references to specific locations or countries in a context

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

Cause

-0.66

INT

-0.66

ronic

-0.63

 eternity

-0.62

zes

-0.60

manac

-0.59

]]

-0.59

asons

-0.59

stars

-0.58

iously

-0.58

POSITIVE LOGITS

 meanwhile

1.25

 however

0.95

 where

0.69

where

0.68

 moreover

0.66

 though

0.65

Vie

0.65

ahime

0.65

 nicknamed

0.63

 Leone

0.62

Activations Density 0.330%

countries or states in the format "In [location]," followed by a significant event or situation

references to specific locations or countries in a context

No Comments

No Known Activations