INDEX

Explanations

keywords related to technical or computer terms, particularly regarding terminals or command lines

oai_token-act-pair · gpt-3.5-turbo

references to terminals or terminal-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

akening

-0.86

yer

-0.84

itsch

-0.83

resh

-0.77

ered

-0.76

sky

-0.76

uph

-0.73

pherd

-0.73

anson

-0.71

lesh

-0.71

POSITIVE LOGITS

 emulator

1.16

 terminal

1.15

 terminals

1.00

 Terminal

0.93

 velocity

0.73

TY

0.70

termin

0.69

 checkout

0.67

igmatic

0.66

 emulation

0.65

Activations Density 0.012%

keywords related to technical or computer terms, particularly regarding terminals or command lines

references to terminals or terminal-related concepts

No Comments

No Known Activations