INDEX

Explanations

words related to records and achievements

oai_token-act-pair · gpt-3.5-turbo

terms related to records and breaking records

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ghan

-0.75

lez

-0.69

inki

-0.68

hirt

-0.67

atown

-0.65

afer

-0.63

Ã§

-0.63

esh

-0.62

gaard

-0.62

wagon

-0.62

POSITIVE LOGITS

 rainfall

0.82

 deeds

0.74

 consecutive

0.72

 recorded

0.70

 accomplishment

0.69

record

0.69

recorded

0.67

0.66

breaking

0.66

setting

0.65

Activations Density 0.202%

words related to records and achievements

terms related to records and breaking records

No Comments

No Known Activations