INDEX

Explanations

words related to celebrations and events

oai_token-act-pair · gpt-3.5-turbo

references to specific characters or elements in scripted narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 tremend

-0.74

ernaut

-0.66

 unden

-0.64

 bully

-0.61

PDATE

-0.61

ftime

-0.59

Orderable

-0.59

 Huck

-0.59

ashtra

-0.58

 reluct

-0.57

POSITIVE LOGITS

idis

0.67

ttes

0.66

 corpus

0.65

 Profession

0.64

 Parables

0.64

Initialized

0.64

rette

0.63

Cir

0.61

hyde

0.60

gypt

0.59

Activations Density 0.505%

words related to celebrations and events

references to specific characters or elements in scripted narratives

No Comments

No Known Activations