INDEX

Explanations

instances where something or someone comes very close to a specific point or threshold

oai_token-act-pair · gpt-3.5-turbo

phrases indicating proximity or nearness in context

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

Definition

-0.68

de

-0.66

 goodbye

-0.64

rw

-0.63

FORE

-0.62

dragon

-0.61

ple

-0.61

ãĥ£

-0.60

enders

-0.60

ãĥĹ

-0.60

POSITIVE LOGITS

 bounds

1.00

ciating

0.84

isine

0.83

isode

0.83

ĵĺ

0.79

 reach

0.77

 inches

0.76

 parentheses

0.75

ĺħ

0.75

autical

0.74

Activations Density 0.026%

instances where something or someone comes very close to a specific point or threshold

phrases indicating proximity or nearness in context

No Comments

No Known Activations