INDEX

Explanations

proper nouns and phrases related to titles or named entities

oai_token-act-pair · gpt-3.5-turbo

variations of the verb "to bed" and related forms indicating a state or condition

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

adra

-0.82

SPONSORED

-0.80

apo

-0.74

pta

-0.73

feat

-0.71

Thor

-0.69

TYPE

-0.68

ede

-0.68

endi

-0.68

antage

-0.67

POSITIVE LOGITS

 aback

0.93

 Leader

0.75

 Victims

0.70

 Thoughts

0.69

 Advertisement

0.67

By

0.67

 Loaded

0.66

 Procedures

0.66

 Benefits

0.65

 Tracks

0.63

Activations Density 0.104%

proper nouns and phrases related to titles or named entities

variations of the verb "to bed" and related forms indicating a state or condition

No Comments

No Known Activations