INDEX

Explanations

terms related to active involvement or engagement in various contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to participation in various contexts

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

otropic

-0.68

enegger

-0.66

 paraph

-0.65

enthal

-0.65

ASC

-0.64

ricanes

-0.64

eric

-0.63

Sensor

-0.63

ICAL

-0.61

pat

-0.61

POSITIVE LOGITS

 participation

0.98

 Participation

0.85

idable

0.81

 rate

0.80

 rates

0.76

 Rate

0.75

rates

0.74

rate

0.72

 testers

0.71

avement

0.71

Activations Density 0.028%

terms related to active involvement or engagement in various contexts

terms related to participation in various contexts

No Comments

No Known Activations