INDEX

Explanations

instances of the word "came", particularly in the context of people entering a location

oai_token-act-pair · gpt-3.5-turbo

repetitions of the word "came."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

²¾

-0.92

olor

-0.84

illusion

-0.82

inguished

-0.76

hedon

-0.74

orthodox

-0.73

relevant

-0.73

elsius

-0.72

raid

-0.72

guided

-0.70

POSITIVE LOGITS

 undone

1.12

 ashore

0.93

 forth

0.82

 pouring

0.80

out

0.79

 flooding

0.74

 crashing

0.73

up

0.73

 closer

0.72

 forward

0.71

Activations Density 0.057%

instances of the word "came", particularly in the context of people entering a location

repetitions of the word "came."

No Comments

No Known Activations