INDEX

Explanations

the word "long" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to long-term concepts or durations

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

IRO

-0.77

ILA

-0.74

atche

-0.71

leck

-0.69

illard

-0.69

gency

-0.67

ropolitan

-0.67

ashtra

-0.66

illery

-0.64

ECH

-0.64

POSITIVE LOGITS

itud

1.14

sword

1.05

lasting

0.98

 overdue

0.98

ago

0.97

 enough

0.96

itudinal

0.95

leaf

0.95

 lasting

0.94

itude

0.92

Activations Density 0.051%

the word "long" in various contexts

references to long-term concepts or durations

No Comments

No Known Activations