INDEX

Explanations

long and foreign-sounding words, potentially related to geographical locations

oai_token-act-pair · gpt-3.5-turbo

patterned language and structures in text

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

 corrid

-0.82

 millenn

-0.72

 referen

-0.71

 helicop

-0.68

 conclud

-0.65

 psychiat

-0.65

 challeng

-0.65

PDATE

-0.64

 nodd

-0.62

 suspic

-0.62

POSITIVE LOGITS

Â®

0.72

idae

0.70

phia

0.65

ai

0.64

zyk

0.61

 Mountains

0.60

 Valley

0.60

 County

0.59

aten

0.59

 Township

0.59

Activations Density 0.378%

long and foreign-sounding words, potentially related to geographical locations

patterned language and structures in text

No Comments

No Known Activations