INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

terms related to division or specificity within a context

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

 situational

-0.79

 redundancy

-0.73

 dime

-0.73

 brim

-0.72

 sacrific

-0.67

 amnesty

-0.66

 sovere

-0.66

 sovereignty

-0.66

 runway

-0.65

 loop

-0.65

POSITIVE LOGITS

milo

0.91

vous

0.87

sky

0.82

rey

0.82

lihood

0.81

ukong

0.80

furt

0.80

å§«

0.80

theless

0.80

sson

0.80

Activations Density 0.164%

proper nouns

terms related to division or specificity within a context

No Comments

No Known Activations