INDEX

Explanations

references to points or experience gained in a game-like setting

oai_token-act-pair · gpt-3.5-turbo

terms related to experience points and rewards in gaming

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Cherokee

-0.74

 Afric

-0.70

 Nicarag

-0.69

 schizophren

-0.67

 Willie

-0.65

unci

-0.64

 Canary

-0.63

 Chero

-0.63

cknow

-0.62

 Siri

-0.62

POSITIVE LOGITS

 Gained

1.08

 expended

1.07

 accumulation

1.04

 depletion

1.03

 accumulated

0.99

 accumulate

0.95

uria

0.89

 gained

0.89

flow

0.88

 totals

0.88

Activations Density 0.543%

references to points or experience gained in a game-like setting

terms related to experience points and rewards in gaming

No Comments

No Known Activations