INDEX

Explanations

phrases indicating a specific duration of time past

oai_token-act-pair · gpt-3.5-turbo

conjunctions and phrases indicating time duration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

=(

-0.68

 Mechdragon

-0.65

REP

-0.64

Neg

-0.64

Safe

-0.62

 Collider

-0.58

restrial

-0.58

BILITIES

-0.57

Own

-0.57

Bal

-0.57

POSITIVE LOGITS

ilipp

0.95

months

0.77

undreds

0.74

 quarter

0.73

 epoch

0.73

essee

0.73

dozen

0.72

terday

0.72

 periods

0.72

quarters

0.71

Activations Density 0.101%

phrases indicating a specific duration of time past

conjunctions and phrases indicating time duration

No Comments

No Known Activations