INDEX

Explanations

phrases indicating a sequence of events

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "follows" which typically indicates a description of events or actions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

orc

-0.72

oll

-0.67

olla

-0.66

uras

-0.65

vere

-0.63

usable

-0.62

owered

-0.61

inese

-0.61

rylic

-0.60

orgetown

-0.59

POSITIVE LOGITS

 follows

1.01

bourg

0.84

ĸļ

0.82

 follow

0.76

 Follow

0.76

=#

0.73

 followed

0.72

follow

0.71

 faithfully

0.71

>:

0.68

Activations Density 0.011%

phrases indicating a sequence of events

instances of the phrase "follows" which typically indicates a description of events or actions

No Comments

No Known Activations