INDEX

Explanations

phrases related to historical records or past events

oai_token-act-pair · gpt-3.5-turbo

phrases related to historical narratives or accounts

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

ateurs

-0.76

agers

-0.76

orney

-0.74

aido

-0.73

iffs

-0.72

rative

-0.72

agraph

-0.72

ando

-0.69

achment

-0.67

mented

-0.67

POSITIVE LOGITS

 history

0.76

 intertwined

0.75

 record

0.72

ĸļ

0.71

 Preservation

0.70

 Revision

0.69

 Timeline

0.68

 Lith

0.66

KH

0.66

CTR

0.65

Activations Density 0.030%

phrases related to historical records or past events

phrases related to historical narratives or accounts

No Comments

No Known Activations