INDEX

Explanations

configurable defaults or settings in software or programming code

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "default" along with related numerical values

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

borough

-1.02

atform

-0.93

lehem

-0.84

ropolitan

-0.81

iddles

-0.79

chn

-0.79

andan

-0.78

hemat

-0.75

wick

-0.75

aph

-0.75

POSITIVE LOGITS

 defaults

1.01

 settings

1.00

 behaviour

0.93

 behavior

0.86

 default

0.86

 configuration

0.80

 gateway

0.79

isable

0.77

 implementations

0.76

 dialog

0.73

Activations Density 0.015%

configurable defaults or settings in software or programming code

occurrences of the word "default" along with related numerical values

No Comments

No Known Activations